How do I stop being the bottleneck? A 90-day plan

It is 8.17 on Monday morning. Your sales director wants a price checked. At 8.26, somebody needs your view on a refund. By nine, three competent adults are waiting for you.

If every decision still runs through you, the business hasn’t scaled. You’ve just got busier.

Every time you jump in, fix the issue and save the day, you teach the company that waiting for you is quicker and safer than deciding without you. The rescue makes you feel useful. The lesson it leaves behind is bloody expensive.

I reviewed 30 recorded founder-CEO calls in June 2026. Again and again, I heard the same five words: “Everything goes through me.”

Of course everything goes through you. You have spent years becoming the quickest route through the business. In the early days, that was an advantage. Now it is where the traffic jams.

The company can only move as fast as one person can think.

And in your company, that person is you.

This 90-day plan shows what should change inside a serious coaching engagement. If you are still choosing support, compare it with my guide to the best CEO coaches for scaling UK businesses.

Why do capable teams keep waiting for the founder?

Capable teams wait because the founder has made waiting rational.

Imagine a restaurant where every plate needs the head chef’s approval before it leaves the kitchen. You would not call that exceptional quality control. You would call it a queue. Yet founders create the same arrangement around prices, hires, refunds, proposals and customer complaints, then wonder why lunch never arrives.

Now picture a five-lane motorway forced through one toll booth. It does not matter how fast the cars are. The road has been designed to jam.

The kitchen and the motorway have the same problem. Plenty of capacity, one compulsory stop.

A scaling company can be built the same way. The people may be good and the org chart may look perfectly sensible. But if permission, context and political cover all sit with you, the company has one real operating rule:

Ask the founder.

At Monkhouse & Company, I have coached more than 200 founder-CEOs and supported 12 founder-led exits. I keep seeing the same split. The formal structure says the team owns the work. The real structure says the founder owns the answer.

Now look, I am not suggesting you disappear to play golf while everybody else guesses what you meant. That merely replaces the queue with a guessing game.

Your team needs context and authority together. You keep the accountability visible, but you stop demanding a preview of every answer.

It is harder than saying “just own it”, of course. It also gives them something they can actually own.

Where is the bottleneck actually hiding?

Look at the queue, not the org chart.

Your job description will tell you that the commercial director owns pricing and the customer director owns refunds. Your Monday morning will tell you whether either statement is true.

I call this the queue test. It is gloriously unsophisticated.

For one week, record every time somebody asks you to approve, review, decide, explain or rescue something. Do not improve the system yet. Just catch yourself in the act.

Then sort each request into four groups:

1. Only I can decide it.
2. Somebody else could decide it with clearer limits.
3. The answer should live in a rule or playbook.
4. The work should stop.

The first group should be small. Strategy, capital, the most senior appointments and a handful of existential risks belong there. Standard discounts, routine hires, supplier substitutions and ordinary customer remedies usually do not.

Make the most frequent request in groups two or three your first transfer. Easy transfers may make your list look tidy. Frequent ones change the way work moves.

I have watched smart founders tell a colleague to “own it” while keeping the knowledge, permission and political cover that ownership requires. Nothing moves. The decision simply comes back wearing different clothes.

A proper founder coaching diagnostic should expose where your calendar makes you the approval point, escalation path or emotional pressure valve.

Why does delegation keep failing?

Delegation fails because founders hand over the task and keep the judgement.

You tell somebody to own pricing, then quietly retain the mental model for what a good price looks like. You give away hiring, but keep the unspoken standard for who is good enough. You delegate the customer problem, then reserve the right to change the answer because it does not feel quite right.

It is like handing somebody a map with no legend. They can see the roads, but they cannot tell which one matters. Sensible people ask the cartographer.

If the judgement stays with you, so does the decision.

I use the context test. Can the owner explain the outcome, the limits and the reason behind the decision without asking you? If not, the work has moved but the authority has not.

Until people can borrow your reasoning, you have moved the task and lengthened the journey. The decision still ends up with you.

What should change in the first seven days?

For the first seven days, do not delegate anything new. First, make the invisible queue visible.

For each request, capture four things:

• who asked
• what they needed
• why it came to you
• what would have allowed them to decide without you

The last question does most of the work. If your answer is “they need better judgement”, what does better judgement mean here? A margin threshold? A customer principle? A spending limit? A strategic priority? Confidence that you will back a reasonable call?

“Use better judgement” is what founders say when the standard still lives in their head. If you cannot explain the standard, nobody else can reliably apply it.

By the end of the week, choose one recurring decision and name one owner. Write down the outcome they own, the authority they have and the three conditions that bring the decision back to you.

Keep it to three conditions, not 17. Seventeen exceptions is a polite way of saying you have not delegated it.

What should be true by day 30?

By day 30, one meaningful decision stream should run without you. I have seen clients get a day a week back within 90 days. You earn that day one decision at a time.

Take refunds. A customer lead might approve them up to an agreed value, provided the decision remains inside margin and reputational guardrails. You review the pattern once a week. You do not edit every call before it happens.

I know what you’re thinking. What if they make a worse decision than you would?

Sometimes they will. Not every time, but often enough to irritate you.

This is where founders discover that what they called ownership was really telepathy. They wanted somebody else to make the decision, provided it was exactly their decision, reached by exactly their route, at exactly their speed.

Telepathy is a rotten management system.

Your team is allowed to decide differently. They are not allowed to ignore the agreed limits.

You intervene when the decision breaches the agreed limits, not when it offends your personal preference. If the limits are wrong, improve them in the weekly review. Do not snatch the work back and declare that nobody can be trusted.

This is where CEO coaching earns its keep. A good coach does not simply help you think. They challenge the behaviour that keeps pulling authority back into your hands.

What must happen between days 31 and 90?

Between days 31 and 90, turn that first transfer into an operating habit.

Add a weekly decision review. Look at the calls made, the outcomes produced and the exceptions triggered. Teach the principle once. Change the boundary when the evidence says it is wrong. Then transfer the next recurring decision.

This rhythm beats another heroic burst of delegation. Problems surface before they reach your desk. Information moves without you dragging it through the company. Decisions happen near the people who have the useful knowledge.

Track five numbers:

• approvals waiting for you
• recurring questions
• decisions made without you
• issues escalated outside the agreed limits
• days spent on business-as-usual work

Then run the calendar test. If business as usual still consumes five days, your title may have changed but your job has not.

The destination is the Two-Day-Week CEO Blueprint^™: three days on CEO-only work and two days on BAU. Strategy, capital, senior talent and the market should get most of your week.

This will feel uncomfortable. Busyness has probably become part of the evidence you use to prove that you matter, so a clear patch in the diary can look suspiciously like unemployment.

Leave it clear. Capacity always looks empty before you decide what deserves it.

By day 90, I care less about the tools a coach has introduced than the decisions that no longer need you. A warm feeling after a coaching session is lovely, but it is not an operating result.

The measure is subtraction.

Use the Two-Day-Week CEO Blueprint™ as the destination, not as a slogan.

Why should you trust this advice?

I don’t think I’m that clever. What I do have is repetition. I have built and scaled two firms, then watched the same decision queue appear across hundreds of coaching conversations.

After enough Monday mornings, the queue stops looking mysterious.

Which support model fits the problem?

Choose the lightest support that changes what happens on Monday. Vistage says its peer groups contain 12 to 16 leaders from non-competing businesses. That breadth is useful when you need more perspectives. It will not do the same job as one-to-one coaching or an executive who owns delivery.

Each model moves a different part of the problem. A coach changes how you lead. A fractional COO owns operating delivery. A peer group improves the range and quality of your thinking.

Buy the wrong one and the queue may acquire nicer language without getting any shorter.

Compare the 1:1 route with CEO peer advisory before you buy the wrong help.

How do you know the coaching is working?

Coaching is working when better decisions happen without you in the room. Agree the baseline at the start and review it at days 30, 60 and 90. You want fewer approvals, faster decisions, fewer repeated escalations and more CEO-only time.

I ask one blunt question at day 90: what can the company do now that it could not do without you three months ago?

If the answer is “nothing”, challenge the diagnosis. End the engagement when you can sustain the new behaviour alone, or when the evidence says you need different help.

The subtraction test is deliberately unsentimental. The queue should be shorter and your calendar should have changed. Otherwise you have hired another meeting. You already had enough of those.

Frequently asked questions

How do I know if I’m the bottleneck in my business?

Look for ordinary decisions waiting in your inbox, projects that stall when you are away and capable people bringing you questions their roles should answer. The issue is not your personality. The company still depends on context and authority that live with you.

Why do founders become bottlenecks without meaning to?

The business was built around the founder’s judgement, and the habit survives long after the early stage has passed. When decision rights are vague and nobody has agreed what good looks like, asking the founder is the sensible response. The queue grows even if the founder has no particular hunger for control.

What’s the difference between delegating tasks and delegating judgement?

Delegating a task gives somebody the work. Delegating judgement gives them the outcome, the limits and enough reasoning to make the call without you. If they still need your blessing before acting, the task has moved but the decision has not.

Which decisions should only the founder-CEO make?

Keep strategy, capital allocation, the most senior appointments and decisions that could threaten the company. Put repeatable and reversible decisions with the people closest to the work. A question that reaches you every week needs a named owner or a rule.

What if my team makes a worse decision than I would?

Sometimes they will. Ask whether the call stayed inside the agreed limits and produced an acceptable result. Review the pattern and improve the boundary. Take the work back after one imperfect decision and everyone learns that ownership lasts only until you disagree.

How quickly should I know whether the bottleneck is moving?

Within 30 days, one meaningful decision stream should run without you. By day 90, the approval queue and your calendar should both look different. I have seen clients regain a day a week within 90 days. If nothing has left your plate, something is wrong with the work.

Start with one decision

Start with last week, not a personality test. Write down every decision that waited for you. Circle the one that appeared most often. Name the person who should own it, the result they own and the three reasons it should return to you.

I want you to run that transfer for 30 days before you buy another productivity system.

The company does not need to learn to live without you. It needs to learn to decide without you.

If you cannot let one ordinary decision go, you have not built a company that can ever let you go.